基于GPS/INS紧组合导航的UKF和CKF滤波精度分析  

Cubature and Unscented Kalman Filter Accuracy Analysis Based on Tightly-Coupled GPS /INS Integrated Navigation System

作　　者：孟 朱立新[1] 

机构地区：[1]电子工程学院电子制约技术安徽省重点实验室

出　　处：《航天控制》2014年第1期10-15,20,共7页Aerospace Control

摘　　要：GPS/INS组合导航系统采用以误差量作为状态变量的非线性紧组合模型,通过函数泰勒展开式,理论分析了无迹卡尔曼滤波(UKF)和容积卡尔曼滤波(CKF)2种算法的滤波精度,引入了数值稳定因子,并证明当状态维数高于三维时,UKF积分精度变差,而CKF的滤波精度高于UKF。将2种算法应用于组合导航中,通过仿真实验,验证了理论分析的正确性,所得结果具有工程应用参考价值。The using non-linear and tightly-coupled model is used in GPS/INS integrated navigation system,which is based on errors as state variables.Analysis accuracy of unscented Kalman filter （UKF） and cubature Kalman filter （CKF） from the Taylor expansion of the function and stability factor is introduced to prove that the accuracy of UKF is deteriorated and the filtering accuracy of CKF is better than UKF when the state dimension is more than three dimensions.Two algorithms are used in integrated navigation system and simulation results verify the correctness of the theoretical analysis and offer certain reference value for engineering applications.

关 键 词：GPS INS组合导航 UKF CKF 泰勒展开式 稳定因子 导航精度 

分 类 号：TN967.2[电子电信—信号与信息处理]